Version 1.77.0 breaks evaluation of custom functions

[TYsewyn]

Version 1.77.0 breaks runtime evaluation of custom functions with multiple arguments.

Assume following ABAC model:

[request_definition]
r = sub, obj, act

[policy_definition]
p = sub_rule, obj, act

[policy_effect]
e = some(where (p.eft == allow))

[matchers]
m = eval(p.sub_rule) && r.obj == p.obj && r.act == p.act

When we register a custom function named hasNameAndAge and add a policy to the enforcer as follows, then the enforcer throws runtime exceptions. In this case Execute 'eval' function error, nested exception is: EOF while reading string at index: 20

@Test
public void testEvalCustomFunction() {
    Enforcer e = new Enforcer("examples/abac_rule_model.conf", "examples/abac_rule_policy.csv");
    HasNameAndAge customFunction = new HasNameAndAge();
    e.addFunction(customFunction.getName(), customFunction);

    List<String> rule = new ArrayList<>();
    rule.add("hasNameAndAge('alice,green', 18)");
    rule.add("data1");
    rule.add("read");
    e.addPolicy(rule);

    TestEvalRule aliceGreen = new TestEvalRule("alice,green", 18);
    testEnforce(e, aliceGreen, "data1", "read", true);
}

public static class HasNameAndAge extends CustomFunction  {

    @Override
    public String getName() {
        return "hasNameAndAge";
    }

    @Override
    public AviatorObject call(Map<String, Object> env, AviatorObject arg1, AviatorObject arg2) {
        String name = FunctionUtils.getStringValue(arg1, env);
        Number age = FunctionUtils.getNumberValue(arg2, env);
        TestEvalRule sub = (TestEvalRule) env.get("r_sub");
        return name.equals(sub.getName()) && age.intValue() == sub.getAge() ? AviatorBoolean.TRUE : AviatorBoolean.FALSE;
    }
}

Although this is a rather simple custom function and example, it demonstrates a major flaw in the change that was introduced with this commit, specifically the following snippet:

if (hasEval) {
    pvals = splitCommaDelimitedList(pvals);
}

[hsluoyz]

@TYsewyn this bug is introduced in this PR: #443 by @tx2002

@TYsewyn can you make a PR to fix it?

[TYsewyn]

IMO the solution to this bug is to roll the previous change back and look at the adapter.
See #440 (comment)

[hsluoyz]

@TYsewyn PR: #443 has been reverted by PR: #448